Electrical Grid & Substation Resilience

The problem: protection can't be eventually consistent

Grid protection and switching are not analytics workloads — they are control-loop decisions with a latency budget measured in milliseconds, because the cost of a late decision is a cascading fault, not a stale dashboard. A protection scheme that depends on reaching a distant control center before it can act has already failed at the one moment it matters most: when the link to that control center is degraded or down, which is disproportionately likely during the same storm or fault condition that triggered the protection event in the first place.

Distributed generation and renewables make this harder, not easier. A grid with thousands of small inverters and distributed resources needs local, fast coordination between substations and field devices, not a single centralized brain that every decision has to route through.

How Relay fits

• →Deterministic, sub-millisecond latency: protection and switching logic running on Relay at the substation gets the same latency guarantee regardless of what the wide-area network is doing.
• →Continues through connectivity loss: Relay never halts message flow or disables an enabled capability because a control-plane connection or licence check couldn't complete — the substation keeps protecting itself.
• →Native SNMP and industrial protocol support: substation automation and RTUs connect directly, without a translation gateway between a protection relay and the logic watching it.
• →Tamper-evident audit log: every switching and protection action is logged for the post-event review that a regulator or a NERC CIP-aligned compliance program requires.

A typical architecture

Protection relays, RTUs and distributed-generation inverters publish telemetry and status to a Relay cluster running at the substation itself, not in a regional data center. Local protection and switching logic subscribes to that stream and acts within the substation's own latency budget. A filtered, aggregated view is forwarded to the utility's control center for situational awareness and long-term analysis, but the substation's own protection decisions never wait on that forward path — they are already made and logged locally by the time anything reaches the control room.

Compliance as a deployment property, not an add-on

Because Relay runs identically whether deployed on a utility's own hardware or air-gapped inside a critical-infrastructure enclave, a grid operator can satisfy the same architectural requirements — no standing cloud dependency, break-glass-only diagnostic access, FIPS mode where required — that OT security review for bulk electric system infrastructure demands, without adopting a different platform for the substation than for the rest of the business.